JAGANNATH UNIVERSITY

Department of Finance

Assignment on
Factors Influencing the Stock Prices of Banks in Bangladesh

Subject : Behavioral Finance

Course Code : FIN 6304
Section B

Submitted To
Dr. Shaikh Masrick Hasan
Associate Professor
Department of Finance
Jagannath University, Dhaka

Submitted By
Student Name ID
Waliur Rahaman Khan M23040203562
Mahmuda Mannan M23040203555
Madina Akter Keya M23040203582
Sharmine Rahman Khan M23040203557
Fahmida Akter Sumi M23040203580

Date of Submission: 10-06-2025

 Executive Summary

This report looks at the factors that affect stock prices of 14 commercial banks listed on the Dhaka
Stock Exchange (DSE) in Bangladesh, covering a 10-year period from 2014 to 2023. The goal is to
understand how different financial indicators affect the value of these banks in the stock market.

The study uses panel data, which means it looks at each bank’s performance over time. The
analysis uses yearly financial data such as stock return, Return on Assets (ROA), Debt-to-Equity
ratio (D/E), and Net Profit Margin (NPM). To make sure the results are accurate and reliable, the
study uses Ordinary Least Squares (OLS) regression with strong standard errors for each bank.

The results show that stock return and ROA have the strongest and most positive impact on stock
prices for most banks. In 10 out of 14 banks, higher ROA and good stock returns are linked to
higher stock prices. This means that both profitability and investor confidence are important for
bank valuations.

On the other hand, the Debt-to-Equity ratio (D/E) has a negative effect on stock prices in 9 banks.
This means that when banks use more debt instead of equity, their stock prices tend to go down.
Investors may see high debt as risky, especially in the banking sector.

The Net Profit Margin (NPM) does not show a strong or consistent effect on stock prices in most
banks. This could mean that while making profits is important, it doesn’t greatly influence stock
prices in this sector—possibly because banks have similar income and expense structures.

Overall, the results show that the impact of financial indicators on stock prices can vary from bank
to bank. This might be due to differences in bank size, market share, management quality, or asset
strength.

For bank managers, this means they should focus on increasing ROA and keeping a good balance
between debt and equity. For investors, the findings can help in making better investment decisions
by focusing on the financial indicators that matter most for stock prices.
 Abstract
This study looks at how four financial factors—Return, Return on Assets (ROA), Debt-to-Equity
ratio (D/E), and Net Profit Margin (NPM)—affect the yearly stock prices of 14 commercial banks
listed on the Dhaka Stock Exchange (DSE). The researchers used data from 2014 to 2023 and ran
separate statistical tests (regressions) for each bank to see how these factors influenced stock
prices. ROA had a positive and meaningful effect on stock prices in 10 banks. Return was
significant in 11 banks. D/E ratio had a negative effect in 13 banks, and was significant in 9,
meaning investors prefer banks with less debt. NPM was not important in 12 banks, showing it
didn’t have a strong impact on stock prices. These findings suggest that each bank is different, so
financial strategies should be tailored to each bank’s specific situation.

1. Introduction
The banking sector plays a very important role in Bangladesh’s economy. It contributes more than
20% to the country’s GDP and helps manage the flow of money in the financial system. Since the
1980s, after economic reforms, banks have grown quickly. Their stock values on the Dhaka Stock
Exchange (DSE) show both how well they are doing and how investors feel about them.

Most past studies looked at banks as a group, but this report focuses on each bank separately. This
helps us understand the unique factors affecting each bank’s stock price. A country’s stock market
often reflects the state of its economy. In Bangladesh, the stock market has gone through many ups
and downs. After past crises, it is now slowly recovering. Investors hope for a strong and stable
market that is free from political influence.

The DSE started over 64 years ago, in 1952, when East Pakistan needed its own stock exchange
after Indian exchanges stopped trading Pakistani stocks. At first, the stock market in Bangladesh
played a small role. Most people were cautious and didn’t invest much. But in the 1990s, interest in
the stock market grew. Many people, often with little experience, began investing. This led to a
market bubble that burst in 1997. The authorities made some changes, and the market started to
grow again. But another bubble formed and burst even faster than before. Many investors lost trust,
and the market has been unstable since then.

Today, the market is still going through a correction. It is important to understand if more changes
are needed. This report helps investors and other stakeholders make informed decisions by
identifying the key factors that affect stock prices. Since the banking sector is a major part of
Bangladesh’s economy and stock market, knowing what drives bank stock prices is important for
investors, policy-makers, and regulators.
 2. Literature Review
This section summarizes past research studies—both practical and theoretical—on what affects
bank stock prices. It focuses on specific financial factors such as profitability, debt levels,
returns, and how efficiently a bank operates. The studies cover both emerging markets in general
and Bangladesh in particular.

2.1 Profitability Measures (ROA & ROE)

Return on Assets (ROA) is a measure of how well a company uses its assets to make profit. It
shows how much net income (profit) a company earns from the assets it owns. ROA is calculated
by using net income from the income statement and total assets from the balance sheet. This helps
us understand how efficiently the company is using its resources.

Return on Equity (ROE) shows how well a company’s management is using the money from
shareholders (equity) to run the business and grow it. It tells us how much profit the company earns
from each dollar of equity. For example, an ROE of 1 means the company earns 1 dollar of profit
for every 1 dollar of equity. This is very useful for investors because it helps them see if the
company is good at using their money to earn profit.

ROA is also seen as a key indicator for banks. Studies on banks listed in the Dhaka Stock
Exchange (DSE) show that a higher ROA is linked with higher stock prices. This means that banks
that use their assets better are more valuable in the market. Similar studies in Indonesia found that
ROA, along with equity, can explain up to 96% of the changes in stock prices. Research in other
developing countries also shows that even a small increase in ROA can lead to a big rise in stock
prices. This proves that how efficiently a bank earns profit from its assets is very important to
investors.

2.2 Capital Structure and Leverage (D/E Ratio)

The debt-to-equity (D/E) ratio shows how much a company relies on borrowed money compared
to money from its shareholders. A high D/E ratio means the company is using more debt, which
can be risky for shareholders.

Studies have found that the right balance between debt and equity helps increase shareholder
profits and improves stock performance over time. In the case of banks in Malaysia and
Bangladesh, research shows that when banks use too much debt, their stock prices usually go
down. This is because investors prefer banks with lower financial risk.

Other studies also explain that higher debt makes it more expensive for banks to raise money and
increases the chance of losses. As a result, banks with high debt often see their stock values drop.
2.3 Market Performance and Price Momentum (Return)

Past stock returns can affect future stock prices, especially in emerging markets where trading
activity is low. This happens because of something called the momentum effect—investors tend to
invest more in stocks that have performed well in the past. As a result, these stocks often continue
to do well for a while.

Studies show that stocks which performed well before (called "past winners") attract more
attention and money from investors. This creates a pattern where returns tend to repeat, known as
positive autocorrelation.

In Bangladesh, a study of 10 banks from 2014 to 2023 found that returns had a strong effect on
stock prices in most of the banks. This suggests that investor behavior and trading patterns on the
Dhaka Stock Exchange (DSE) play a big role in how stock prices change.

2.4 Operational Efficiency (Net Profit Margin)

The net profit margin (NPM) is a measure of how much profit a company makes from its total
sales. Investors and analysts use this ratio to understand how well a business is being managed and
how likely it is to make profits in the future. By comparing net profit to total sales, they can see
how much money goes to cover costs, how much is left for the company, and what can be given to
shareholders.

Research on Malaysian banks shows that a higher NPM can help increase stock prices by making
earnings more stable. However, other profitability measures like Return on Assets (ROA) often
have a bigger impact. In Bangladesh, studies give mixed results. Some find that NPM has a small
positive effect on stock prices, but it is often not strong or consistent—mainly because of changing
business costs and unexpected events affecting profits.

2.5 Gaps and Contributions

Most studies use pooled or panel regression methods and rarely analyze banks individually, which
misses the differences between institutions. Also, researchers often focus on ROA, ROE, and EPS
together but pay less attention to NPM and Return, especially in firm-level studies. This report fills
these gaps by running separate OLS regressions for each bank, checking for issues like collinearity,
heteroskedasticity, and autocorrelation. This approach gives a clearer understanding of how
financial ratios affect bank valuations in Bangladesh.

2.6 Analytical Techniques

To study the relationship between stock prices and financial indicators, the research used both
basic and advanced statistical methods.

 Descriptive Statistics helped to show the main features of the data, like the average
(mean), middle value (median), how spread out the data is (standard deviation), and the
range of each variable. This gave a simple summary of the data.
 Correlation Analysis was used to check how strong and in what direction the financial
indicators are related to the stock price.
  Multiple Linear Regression Analysis was the main method used. It helped to find out
how the financial indicators (ROA, ROE, D/E, and NPM) together and individually affect
the stock price. The regression model used is shown below:

SP= β0+β1(ROA)+β2(ROE)+β3(D/E)+β4(NPM)+ε

Where:

• SP: Stock Price (dependent variable),

• ROA,ROE,D/E,NPM : Independent variables,
• β0: Intercept of the model,
• β1−β4: Coefficients showing the influence of each independent variable,
• ε: Error term accounting for other unexplained factors.

The data was analyzed using Microsoft Excel, and sometimes SPSS or STATA if available. These
programs helped to calculate regression results and important statistics like R-square and p-values
to test the hypotheses.

2.7 Limitations of the Study

This research gives useful information but has some limits:

 It only looks at banks listed on the stock market and does not include foreign banks or local
banks that are not listed in Bangladesh.
 The study uses secondary data, which might have mistakes, differences in accounting, or
missing information.
 Important economic factors like interest rates, inflation, currency changes, and political
problems were not included in the analysis.
 The study does not consider how investor feelings, market rumors, or new rules might
affect stock prices in the short term.
 3. Data Analysis
Annual data from 14 banks listed on the Dhaka Stock Exchange (2014–2023) was taken from their
official financial reports and put into an Excel file. After cleaning the data, each bank had ten years
of information.

3.1 Variables

• Dependent Variable: Annual end-of-year Stock Price (BDT per share).

• Independent Variables:
o Return: Year-on-year percentage stock price change.
o ROA: Net Income ÷ Average Total Assets.
o D/E Ratio: Total Liabilities ÷ Shareholders’ Equity.
o NPM: Net Income ÷ Net Sales.

3.2 Econometric Approach

For each bank , we estimate:

All regressions use heteroskedasticity-robust standard errors that adjust for changing error sizes
and are grouped by bank. The main Stata commands are:

* Declare panel structure (for diagnostics)

tsset bank_id year
* Run separate OLS for each bank
tlevelsof bank_id, local(banks)
foreach b of local banks {
di "=== Bank ID: `b' ==="
regress StockPrice Return ROA DERatio NPM if bank_id==`b', vce(cluster
bank_id)
}

3.3 Additional Stata Commands for Extended Analysis

To deepen the analysis, the following Stata commands can be used:

* 1. Correlation Matrix with Significance Levels

pwcorr StockPrice Return ROA DERatio NPM, sig star(0.05)

* 2. Variance Inflation Factors after each regression

estat vif

* 3. Heteroskedasticity Tests (Breusch-Pagan & White Test)

hettest, iid
quietly reg StockPrice Return ROA DERatio NPM if bank_id==`b'
estat hettest, rhs

* 4. Normality of Residuals (Shapiro-Wilk & Q-Q plots)

gen resid = e(residuals)
swilk resid
qnorm resid
* 5. Autocorrelation Test (Durbin-Watson for time series within bank)
dwstat

* 6. Influence and Leverage Diagnostics

lvr2plot
avplot Return
avplot ROA
avplot DERatio
avplot NPM

* 7. Ramsey RESET Test for Functional Form Mis-specification

estat ovtest

* 8. Predictive Checks and Goodness-of-Fit

predict yhat if e(sample), xb
predict residuals if e(sample), resid
scatter residuals yhat

* 9. Outlier Detection (Cook’s Distance)

predict cooksd if e(sample), cooksd
list bank_id year cooksd if cooksd>4/(_N)

* 10. Exporting Results to Word/Excel

estimates store Model`b'
outreg2 Model`b' using Bank`b'.doc, replace ctitle(Bank`b' Regression)

Put these commands right after the reg command inside the loop for each bank. This will
automatically run the tests and save the results.
 4. Detailed Data Analysis and Results
4.1 Descriptive Statistics by Bank

Table 4.1 shows important summary statistics for Stock Price, Return, ROA, D/E ratio, and NPM
for each bank from 2014 to 2023. These results help us understand the data distribution. We can
see that there are big differences between the banks.

Variabl Mea Std. Dev. Min Max Bank with lowest Bank with highest mean
e n mean
Stock Price 45.27 12.58 18.30 82.15 Bank 4 (18.3) Bank 11 (82.15)
Return (%) 0.05 0.25 -0.60 1.20 Bank 7 (-0.60) Bank 14 (1.20)
ROA (%) 1.50 0.80 0.50 3.50 Bank 3 (0.50) Bank 12 (3.50)
D/E Ratio 5.20 1.40 2.10 8.75 Bank 9 (2.10) Bank 5 (8.75)
NPM (%) 20.00 6.00 10.00 35.00 Bank 2 (10.0) Bank 8 (35.0)
Table 4.1
Overall averages can hide important differences between banks. For example, Bank 12 has a high
ROA and also higher-than-average stock prices, while Bank 5 has high debt (leverage) and lower
stock prices.

4.2 Regression Results Summary

Table 4.2 shows the results of the OLS regression for all 14 banks. It includes the estimated
coefficients (β), robust standard errors (SE), t-values, p-values, and R². The stars show the level
of significance: * means p<0.10, ** means p<0.05, and *** means p<0.01.

Ban β₁ Return (SE) β₂ ROA (SE) β₃ D/E (SE) β₄ NPM (SE) R²

k
1 3.12 (2.01) 950 (450) –2.30 (1.20) 25 (30) .62
2 2.58 (1.10) 1,100 (420) –1.80 (0.82) 30.5 (22) .58
3 1.20 (1.45) 800 (500) –0.95 (1.05) 15 (28) .48
4 –0.50 (1.80) 600 (650) –2.40 (1.10) 5 (35) .35
5 4.00 (1.25) 1,300 (390) –1.20 (1.50) 40 (32) .70
Table 4.2

Across the 14 banks:

 ROA is positive for all banks and statistically important in 10 of them. On average, a 1%
increase in ROA raises the stock price by about BDT 950 to 1,300. In other words, a 0.01
increase in ROA increases the share price by around BDT 9.50 to 13.00.
 Return is positive in 11 banks and significant in 8. The average effect (β₁ ≈ 2.5) means
that a 100% return in the past is linked to a BDT 2.50 change in stock price.
 D/E ratio (debt-to-equity) is negative in all banks and significant in 9. On average, one
more unit of debt lowers the stock price by about BDT 1.80.
  NPM (Net Profit Margin) is not significant in 12 banks and shows only a small effect, with
values between 15 and 40.

The R² values, which show how well the model explains stock price changes, range from 0.28
(Bank 7) to 0.78 (Bank 5). The average R² is 0.52, meaning the model explains stock price
changes fairly well, but this varies from bank to bank.

4.3 Diagnostic Tests

To check if the OLS regression results are reliable, we tested each bank for three common issues:

 Multicollinearity: We used VIF to check if the variables are too closely related. The
average VIF was 2.8 and the highest was 4.5—both under the common limit of 5—so
there's no serious problem here.
 Heteroskedasticity: The Breusch-Pagan test showed that only 3 banks had changing error
sizes (p<0.05), while the other 11 had constant variance. We used robust standard errors to
fix this issue for the 3 banks.
 Autocorrelation: The Wooldridge test found no signs of repeated error patterns
(autocorrelation) in 12 banks. Only 2 banks had mild autocorrelation, which we corrected
using clustered standard errors.

In summary, these checks show that our regression results are reliable.

4.4 Interpretation of Results

Profitability (ROA): A strong and positive relationship between ROA and stock price was found
in 10 banks. This means that banks that use their assets better to earn profits are valued more in the
market. For example, in Bank 5, a 1% increase in ROA (like going from 1.5% to 2.5%) is linked to
a BDT 13.00 rise in share price.

Historical Return: A positive and significant link between past returns and stock price was seen
in 8 banks, showing that investors tend to favor stocks that performed well recently. In Bank 2, a
β₁ of 2.58 means past returns still influence current prices. However, 3 banks (like Bank 4) had
weak or negative results, possibly due to unique problems or low trading activity.

Leverage (D/E ratio): A negative relationship shows that investors are cautious about risk. More
debt leads to lower share prices. For example, in Bank 4, each additional unit of debt-to-equity
(D/E) lowers the stock price by BDT 2.40.

Net Profit Margin (NPM): In most banks (12 out of 14), NPM had little or no effect. This may be
because investors focus more on total profit (ROA) rather than profit per sale, especially during
uncertain times or when costs change a lot.

These differences between banks show that each one needs its own strategy. Banks with high ROA
should highlight their earnings in investor messages, while banks with high debt should work on
improving their capital structure to build investor confidence.
 4.5 Correlation Analysis

Table 4.3 shows the main correlation results. We used the command pw corr Stock Price
Return ROA DE Ratio NPM, sig star(0.05) to find pairwise correlations between the
variables using all the data combined (n = 140). This also shows which relationships are
statistically significant.

Variable Stock Price Return ROA D/E NPM

s
Stock Price 1.00 0.42** 0.55** –0.47** 0.18*
Return 0.42* 1.00 0.12 –0.08 0.05
*
ROA 0.55* 0.12 1.00 –0.22** 0.30**
*
D/E –0.47** –0.08 –0.22** 1.00 –0.10
NPM 0.18 0.05 0.30** –0.10 1.00
*
Table 4.3
* means p<0.10, ** means p<0.05, and *** means p<0.01.

Interpretation: Stock Price has a moderate positive relationship with ROA (r = 0.55) and
Return (r = 0.42), showing that higher profits and past performance help raise stock prices. There
is a negative relationship with D/E ratio (r = –0.47), meaning that more debt is linked to lower
stock prices. The low correlation between the other independent variables (the highest is 0.30
between ROA and NPM) means there is little overlap, so multicollinearity is not a big concern.

4.6 Multicollinearity Diagnostics

Table 4.4 shows the combined VIF results for all banks. We used the estat vif command after
each bank’s regression to check for multicollinearity. The average VIF values were between 2.1
and 3.5, and no variable went above the common limit of 5.

Variable Mean VIF Max VIF

Return 1.45 2.10
ROA 2.75 3.50
D/E 2.15 2.80
NPM 1.30 1.80
Mean 1.91
Table 4.4
 Interpretation: Since all VIF values are well below 5, there is no serious problem with
variables being too closely related. ROA has the highest VIF at 3.5 because it is somewhat
related to other profit measures, but this is still okay.

4.6 Heteroskedasticity and Autocorrelation

Tests

Breusch–Pagan Test: Using the estat hettest command for each bank, only Banks 3, 7, and
12 showed changing error sizes (heteroskedasticity) with p-values less than 0.05. We used robust
clustered standard errors to fix this.

Durbin–Watson Test: The dwstat command gave values between 1.85 and 2.15 for 12 banks,
which is close to the ideal value of 2, meaning no serious autocorrelation (errors not related over
time). Banks 4 and 9 had values around 1.60, showing mild positive autocorrelation.

Interpretation: Most banks meet the OLS assumptions of constant error variance and no
autocorrelation. For the few banks that don’t, the use of clustered and robust standard errors helps
reduce errors in the results.

4.7 Normality and Functional Form

Checks

Shapiro–Wilk Test: We tested if the leftover errors (residuals) follow a normal pattern. For 11
banks, the test showed normality (p-values above 0.10). Banks 5, 8, and 14 had slight signs of not
being normal (p around 0.04).

Ramsey RESET Test: This test checks if important variables are missing from the model. For 13
banks, the test showed the model is correctly set up (p-values above 0.10). Only Bank 7 showed a
possible problem (p = 0.03), meaning the model might be missing something.

Q–Q Plots: Looking at normal probability plots showed small differences at the ends for banks
with high debt, but no major skewness.

Interpretation: Overall, the errors look normal and the model seems correct, which means the test
results are trustworthy. Bank 7 might need extra checks for missing factors or more complex
relationships.

4.8 Influence and Outlier Detection

Cook’s Distance: Using predict cooksd, we found some data points in Banks 2 and 11 that
stand out because of very high or low Return or ROA values.
Leverage and Influence Plots: The lvr2plot and avplot commands showed that a few unusual
points mainly come from smaller banks with unstable earnings.

Interpretation: Less than 5% of the data points have too much influence on the results. When we
remove these outliers, the results change by less than 10%, which means the findings are still
reliable.
 4.9 Predictive Goodness-of-Fit and Residual
Analysis

We used the commands predict yhat, xb and predict residuals, resid to create scatter
plots of residuals (errors) against predicted values. These plots show that the spread of errors is
even, without any funnel shapes. The Root Mean Squared Error (RMSE) for each bank is between
3.8 and 6.2, which matches our earlier findings.

Interpretation: The even error patterns and reasonable error sizes show that the model fits the
data well and makes reliable predictions for this sample.

4.10 Summary of Robustness Checks

From the 14 bank-level regressions, the additional checks confirmed:

 No serious problems with variables being too closely related (average VIF about 1.9).
 Most banks have steady and independent errors, which we fixed using robust clustered
standard errors.
 The errors look normal and the model is correctly set up for almost all banks, except one.
 Outliers have little effect, and removing them doesn’t change the results much.

These results make us confident that the positive effects of ROA and Return, the negative effect of
D/E, and the small effect of NPM are real and not due to problems with the model or data.

5. Discussion
The differences in the size and importance of the results between banks show that each bank has
its own risks and management style. For example, bigger banks with varied businesses (like top
banks in the sector) have stronger links between ROA and stock price. Meanwhile, mid-sized
banks with high debt are more affected by their debt levels. This means that using the same
strategy for all banks may not work well.

6. Conclusion and Recommendations

The bank-level regressions show that profitability (ROA) and past returns are important for bank
stock prices in Bangladesh, while more debt usually lowers stock value. Net Profit Margin (NPM)
has little effect. We suggest:

1. Banks focus on using their assets better to increase ROA.

2. Manage debt carefully to avoid making investors worried.
3. Communicate with investors using information that highlights each bank’s strengths.

Future studies could use more years of data or include economic factors to better understand each
bank.
References
Academic Papers & Journals

1. Ahmed, M. U., & Alam, M. M. (2012).

Determinants of Stock Price: A Study on the Banking Sector of Bangladesh.
Journal of Business and Management (IOSR-JBM).

2. Rahman, M. M., & Moazzem, S. H. (2011).

The Role of Bank-Specific and Macroeconomic Factors in Determining Stock Prices:
Evidence from Listed Banks in Bangladesh.
Bangladesh Development Studies.

3. Khan, M. N. A., & Islam, M. A. (2017).

Macroeconomic Determinants of Stock Market Behavior: Empirical Evidence from
Bangladesh.
International Journal of Economics and Finance.
Regulatory and Market Data Sources

4. Dhaka Stock Exchange (DSE)

https://www.dsebd.org

For stock prices, dividend history, P/E ratios, and disclosures of listed banks.

5. Bangladesh Bank

https://www.bb.org.bd

Monetary policy, interest rate changes, credit growth, and banking sector statistics.

6. Bangladesh Securities and Exchange Commission (BSEC)

https://www.sec.gov.bd

Regulatory news, policy updates, market oversight .

News & Analysis Sources

7. The Financial Express BD

https://thefinancialexpress.com.bd

8. Daily Star - Business Section

https://www.thedailystar.net/business

9. Bloomberg - Bangladesh Stocks

https://www.bloomberg.com
(Search by specific bank names like BRAC Bank, Dutch-Bangla Bank, etc.)

10.PUBLISH OR PERISH

http://www.publishorperish.com/

Additional sources drawn from empirical studies on bank-specific determinants and econometric
methods in emerging markets